Scrutable

headshot wrote: ↑

Tue Jan 26, 2021 11:29 am

lpm wrote: ↑

Tue Jan 26, 2021 11:08 am

Aren't there Star Trek episodes and the like where conflict due to limited medicine is the basic plot line?

There's one episode (Symbiosis) where a population is convinced that they'll die if they don't receive life saving medicines supplied exclusively by a neighbouring planet, but it turns out the neighbouring planet has in fact hooked the users on a highly addictive drug with terrible withdrawal symptoms. Picard sorts it all out...by doing nothing.

https://en.wikipedia.org/wiki/Symbiosis ... eneration)

Grumble wrote: ↑

Tue Jan 26, 2021 7:28 am

So Handelsblatt and Bild have based their story on precisely nothing.

The German health ministry has challenged Handelsblatt’s report, saying it mixed up the efficacy rate in over-65s and the proportion of older people involved in the vaccine trials. The German health minister, Jens Spahn, described the Handelsblatt article as “speculation” on Tuesday morning and declined to comment while the European Medicines Agency (EMA) said analysis of AstraZeneca’s trial data was ongoing.

At present, it remains unclear exactly who the source of the German reports is, and what data they might be basing the figure on – making its credibility difficult to assess.

shpalman wrote: ↑

Tue Jan 26, 2021 9:46 am

For all we know, JCVI and AstraZeneca aren't basing their "oh yes it is" on anything either.

In the actual published data we note that the group who accidentally got the LD/SD dose were all 18-55, the efficacy was 90.0%, whereas in the UK SD/SD group it was 60.3% if including the whole age range but 65.6% if limiting to 18-55 to match the LD/SD subgroup.

i.e. COV002 (UK) SD/SD age 18-55 had 34 covid cases of which 8/1407 in the vaccine group and 26/1512 in the control (Table 3).

This would already be a subset of a subset because I note from Table 1 that COV002 (UK; SD/SD) had 1879 and 1922 subjects in the 18-55 range in the vaccine and control groups respectively. The numbers match more closely with the "Health and social care setting workers" (1441 and 1513) and Table 3 does indicate that the models have been adjusted for "BMI (<30 vs ≥30 kg/m2), health-care worker status (yes vs no), and ethnicity (white vs non-white)".

But the COV002 (UK) SD/SD results presented in Table 2 give 15/2377 and 38/2430 cases in vaccine and control groups respectively.

You can't just subtract the Table 3 results from the Table 2 results to get the cases in the >55 age range, because of whatever the f.ck they've done to the data in Table 3, but if you could it would be 7/970 and 12/918.

Because all we can do is make up numbers since we have no idea what data AstraZeneca and the JCVI are looking at.

Gfamily wrote: ↑

Tue Jan 26, 2021 2:40 pm

Grumble wrote: ↑

Tue Jan 26, 2021 7:28 am

So Handelsblatt and Bild have based their story on precisely nothing.

Channel 4 Factcheck

The University of Oxford wrote:“The results of the clinical trials have already been published transparently in five peer-reviewed scientific publications showing similar immune responses in younger and older adults”

shpalman wrote: ↑

Tue Jan 26, 2021 7:23 pm

The French one doesn't work

MV-SARS-CoV-2 was being developed by the Institut Pasteur along with the Coalition for Epidemic Preparedness Innovations, Thémis and MSD.

The Sanofi one is also delayed.

The company will record a pre-tax discontinuation charge in the fourth quarter for vaccine candidate V591, which it acquired with the purchase of Austrian vaccine maker Themis Bioscience, and V590, developed with nonprofit research organization IAVI, Merck said in a statement.

Here it is: the crucial table showing Germany's calculations on the AstraZeneca vaccine's efficacy in over-65s. And good God, look at that confidence interval! For over-65s there was only one infection in the treatment group and one in the placebo group.

Woodchopper wrote: ↑

Thu Jan 28, 2021 2:43 pm

Here it is: the crucial table showing Germany's calculations on the AstraZeneca vaccine's efficacy in over-65s. And good God, look at that confidence interval! For over-65s there was only one infection in the treatment group and one in the placebo group.

https://twitter.com/olivernmoody/status ... 0071860230

In the over 65s there was one infection in the treatment group and one infection in the control group. Which makes the vaccine look ineffective compared to doing nothing.

Looks more like the control group got lucky.

Woodchopper wrote: ↑

Thu Jan 28, 2021 2:43 pm

Here it is: the crucial table showing Germany's calculations on the AstraZeneca vaccine's efficacy in over-65s. And good God, look at that confidence interval! For over-65s there was only one infection in the treatment group and one in the placebo group.

https://twitter.com/olivernmoody/status ... 0071860230

In the over 65s there was one infection in the treatment group and one infection in the control group. Which makes the vaccine look ineffective compared to doing nothing.

Looks more like the control group got lucky.

Sciolus wrote: ↑

Thu Jan 28, 2021 3:30 pm

Or the treatment group got unlucky.

How were the 650 recruited? Most over-65 have been sheltering pretty strictly, unless they're unlucky enough to have to go into hospital or a care home.

bagpuss wrote: ↑

Thu Jan 28, 2021 4:05 pm

Sciolus wrote: ↑

Thu Jan 28, 2021 3:30 pm

Or the treatment group got unlucky.

How were the 650 recruited? Most over-65 have been sheltering pretty strictly, unless they're unlucky enough to have to go into hospital or a care home.

Really (to my bolded bit)? My 85 year old mother has been mostly quite careful but has insisted on going to get her newspaper every day, has continued to see her sister and brother in law, albeit always outside, but definitely against regulations and has continued to have her cleaner once a week and regular hairdresser and chiropodist visits, except during actual lockdowns. A number of my older neighbours have been the most active and out and about of all of us, including some definite rule stretching if not breaking. Going back to the VE Day anniversary celebrations last year, it was notable that everyone under the age of 60 on the street was staying in their own gardens, with only very occasional forays up or down the road to speak to neighbours from a distance, while all the older ones were gathered in one front garden, any initial attempt at some level of distancing completely forgotten after the first beer or wine.

And, of course, the older you are, the more likely you are to have needed some kind of gp or hospital appointment in the last year. Both my mother and father in law have ongoing health conditions which, while not shielding-level serious, have necessitated visits to gp and/or hospital in that time. And in my mother in law's case, hospital appointments, and her vaccine appt, mean being in a car with someone else for between 20 minutes and an hour each way, depending on where they're going, as she doesn't drive.

Sciolus wrote: ↑

Thu Jan 28, 2021 3:30 pm

Or the treatment group got unlucky.

Anecdotally, I do feel like I've seen a lot of "I've been careful and obeyed the lockdown, I only <list of a load of lockdown-breaking things>"

raven wrote: ↑

Thu Jan 28, 2021 5:51 pm

Anecdotally, I do feel like I've seen a lot of "I've been careful and obeyed the lockdown, I only <list of a load of lockdown-breaking things>"

Well, if we're going by anecdote, by way of a counter-example there's my dad. He's been paranoid about getting it, despite being in relatively good health for his age, and has pretty much been a shut-in barring a few essential appointments and a brief trip to meet a new grandkid when we were allowed to do such things. None of us siblings have been inside their house for a year. Not even my sister who lives 10 minutes away and who was delivering shopping regularly in lockdown1, and not even just to access the toilet right by the back door when she & the kids were allowed in the garden...

raven wrote: ↑

Thu Jan 28, 2021 5:51 pm

That said, one case in each arm of a trial group is surely not enough data to draw conclusions from. I'm assuming that was from the UK subjects, over the summer when cases were very low perhaps. Is there no over 65s data from Brasil or South Africa?

Martin Y wrote: ↑

Thu Jan 28, 2021 3:12 pm

Woodchopper wrote: ↑

Thu Jan 28, 2021 2:43 pm

Here it is: the crucial table showing Germany's calculations on the AstraZeneca vaccine's efficacy in over-65s. And good God, look at that confidence interval! For over-65s there was only one infection in the treatment group and one in the placebo group.

https://twitter.com/olivernmoody/status ... 0071860230

In the over 65s there was one infection in the treatment group and one infection in the control group. Which makes the vaccine look ineffective compared to doing nothing.

Looks more like the control group got lucky.

So the efficacy is under 10%. It's exactly 6.3% plus or minus umpty-billion.

Trinucleus wrote: ↑

Thu Jan 28, 2021 6:53 pm

If it's effective in under 65s can you extrapolate how effective it's likely to be for over 65s? Not accurately I appreciate, but enough to be confident about using it?

shpalman wrote: ↑

Tue Jan 26, 2021 2:49 pm

shpalman wrote: ↑

Tue Jan 26, 2021 9:46 am

For all we know, JCVI and AstraZeneca aren't basing their "oh yes it is" on anything either.

In the actual published data we note that the group who accidentally got the LD/SD dose were all 18-55, the efficacy was 90.0%, whereas in the UK SD/SD group it was 60.3% if including the whole age range but 65.6% if limiting to 18-55 to match the LD/SD subgroup.

i.e. COV002 (UK) SD/SD age 18-55 had 34 covid cases of which 8/1407 in the vaccine group and 26/1512 in the control (Table 3).

This would already be a subset of a subset because I note from Table 1 that COV002 (UK; SD/SD) had 1879 and 1922 subjects in the 18-55 range in the vaccine and control groups respectively. The numbers match more closely with the "Health and social care setting workers" (1441 and 1513) and Table 3 does indicate that the models have been adjusted for "BMI (<30 vs ≥30 kg/m2), health-care worker status (yes vs no), and ethnicity (white vs non-white)".

But the COV002 (UK) SD/SD results presented in Table 2 give 15/2377 and 38/2430 cases in vaccine and control groups respectively.

You can't just subtract the Table 3 results from the Table 2 results to get the cases in the >55 age range, because of whatever the f.ck they've done to the data in Table 3, but if you could it would be 7/970 and 12/918.

Because all we can do is make up numbers since we have no idea what data AstraZeneca and the JCVI are looking at.

I think that comes out at about 40%? I don't know, I can't get their exact efficacy percentages from the numbers in the table anymore.

shpalman wrote: ↑

Thu Jan 28, 2021 6:04 pm

raven wrote: ↑

Thu Jan 28, 2021 5:51 pm

Anecdotally, I do feel like I've seen a lot of "I've been careful and obeyed the lockdown, I only <list of a load of lockdown-breaking things>"

Well, if we're going by anecdote, by way of a counter-example there's my dad. He's been paranoid about getting it, despite being in relatively good health for his age, and has pretty much been a shut-in barring a few essential appointments and a brief trip to meet a new grandkid when we were allowed to do such things. None of us siblings have been inside their house for a year. Not even my sister who lives 10 minutes away and who was delivering shopping regularly in lockdown1, and not even just to access the toilet right by the back door when she & the kids were allowed in the garden...

And did they catch less covids, as compared to the "no idea where I could have caught ok here are my contacts for the last two weeks <massive long list of contacts>" anecdoges?

shpalman wrote: ↑

Thu Jan 28, 2021 6:04 pm

The most recent publication is this one: https://www.sciencedirect.com/science/a ... 3620326611

Table 1 indicates that there were >55's in both the UK and Brazil groups, just not the LD/SD UK group. Most of the >70's were in the UK SD/SD group though.